Characterizing the Constitutive Behaviour and Austenite Decomposition of an 1,800 MPa Press Hardenable Boron Steel

Abstract

The mechanical behaviour of a new 1,800 MPa strength grade of Al-Si coated, 37MnB5 press-hardenable steel, designated PHS 1800, was investigated under hot stamping conditions. Prior to mechanical testing, a novel heat treatment technique was developed to pre-alloy the Al-Si coating thereby preventing it from melting during austenitization. The alloy was held at 700°C for 7 – 10 minutes and then quenched to room temperature. This heat treatment technique prevented loss of DIC speckle patterns due to melting of the coating, thereby enabling digital image correlation (DIC) strain measurements of the test specimen at elevated temperatures, and was not observed to affect the constitutive behaviour of the material prior to necking. For constitutive testing, the pre-alloyed material was austenitized and then immediately deformed at strain rates varying between 0.01 and 1s-1 and temperatures between 500°C and 900°C. The experimental tensile test results were processed to obtain the material flow curves, such that the data could be used for finite element simulation of hot stamping processes. An identical series of experiments was also performed on the widely studied 1,500 MPa 22MnB5 hot stamping grade (designated PHS 1500) in order to validate the testing methods and to provide a baseline for assessment of the new PHS 1800 grade. The PHS 1500 results from the current research were compared to several results published in the literature and were found to be in good agreement, indicating the pre-alloying heat treatment and DIC techniques utilized in the current work were valid for the constitutive testing of PHS. Compared to PHS 1500, this new grade of PHS 1800 demonstrated approximately 20% higher flow stress at any given strain level, at the same temperature and strain rate test conditions. This relationship also holds true for both materials after quenching to a fully martensitic state, where the ultimate tensile strength of PHS 1800 is approximately 2,000 MPa and 25% higher than the tensile strength of PHS 1500. For the elevated temperature tensile tests, PHS 1800 exhibited similar strain rate and temperature sensitivities as PHS 1500, wherein higher strain rates or lower temperatures resulted in higher flow stresses. PHS 1800 also experienced necking at up to 31% lower strains than PHS 1500, with the strain to necking varying depending on the test conditions. A modified Norton-Hoff constitutive model, with a second-order polynomial equation to describe the variation in the strain hardening coefficient and strain rate coefficient with respect to temperature, was fit to the experimental results for both PHS 1500 and PHS 1800. An excellent fit (R-squared value of 0.971) was achieved overall at all temperatures and strain rates. A finite element model of the elevated temperature tensile test was created to evaluate the accuracy of the fit curves, and less than 12% error was observed in the predicted load versus displacement response. In addition to the mechanical behaviour of the material, the austenite decomposition characteristics of PHS 1800 were also investigated via constant cooling rate experiments. The effect of deformation on the austenite decomposition was investigated at three strain levels (0%, 10%, 20%), for three different cooling rates (5°C/s, 10°C/s, 50°C/s). It was found that higher cooling rates resulted in greater martensite phase fractions and higher microhardness values. Specimens cooled at 5°C/s and 10°C/s resulted in complex microstructures including ferrite, bainite, and martensite, while the 50°C/s specimens were fully martensitic. The range of hardness results (with no induced deformation) varied between 430HV to 570HV. When deformation was applied to PHS 1800 in its austenitic state, the fraction of ferrite and bainite phases increased substantially after quenching, resulting in a hardness as low as 320HV, a drop of up to 26%. For the specimens cooled at 50°C/s, induced deformation had no effect on the phase fraction or hardness values. Finally, the hardness and phase composition results of the constant cooling rate experiments were utilized to update an existing LS-DYNA PHS material model to enable the prediction of accurate phase change and hardness values for PHS 1800. The updated material model was capable of predicting the final hardness of PHS 1800 to less than 2% error at zero applied deformation, and to less than 14% error for the deformed specimens. The accuracy of the phase fraction predictions varied drastically, ranging from 8% to 56% difference compared to the experimental measurements, depending upon how the material parameters were calibrated.